Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/30487
Our environment has a variety of different features that our visual system must encode every day. These features can be represented as probability distributions. Previous research has shown that our visual system can encode mean, variance and even the distribution shape, both for colour and orientation ensembles (Chetverikov et al. 2016; Chetverikov et al 2017a). Research has also shown that when learning multiple feature distributions (colour and orientation) simultaneously the properties of the irrelevant distribution impaired the encoding of the relevant one (Hansmann-Roth et al, 2018). In this thesis we have revised a previous experiment where observers where to find either the odd color out or an odd orientation line. We increased the learning trials and used only trials in which orientation was the relevant feature. The lines were still coloured to investigate the role of the task- irrelevant feature on distribution learning. Our results are noisier and the distribution learning is impaired compared to the Chetverikov et al (2016) data. Colour seems to influence the precision of the internal representation. Colour does make the search more difficult. This could stem from pop out effects or that the search seems to change from parallel to a serial search, which results in slower RTs.
|Representing orientation ensembles.pdf||1.33 MB||Opinn||Heildartexti||Skoða/Opna|
|Yfirlýsing um meðferð- BS ritgerð.jpg||761.06 kB||Lokaður||Yfirlýsing||JPG|